Hi! This is my status report for contract work done in December 2023.
# Reputation-based bridge distribution
- Continued Tor Browser integration work
https://gitlab.torproject.org/tpo/anti-censorship/team/-/issues/116
- Debugged some issues with the lox distributor deployment
https://gitlab.torproject.org/tpo/tpa/team/-/issues/41406 https://gitlab.torproject.org/tpo/anti-censorship/lox/-/issues/40 https://gitlab.torproject.org/tpo/anti-censorship/lox/-/issues/42 https://gitlab.torproject.org/tpo/anti-censorship/lox/-/issues/43
# Snowflake
- Fixed excessive logs of closed connections at the Snowflake proxy
https://gitlab.torproject.org/tpo/anti-censorship/pluggable-transports/snowf...
- Released version 2.8.1 of Snowflake
- Provided further review of the SQS rendezvous method
https://gitlab.torproject.org/tpo/anti-censorship/pluggable-transports/snowf...
# Domain Fronting Alternatives
- Looked for cloud providers that offer domain fronting
- Looked for good potential front domains for working cloud providers
In December, I started a short contract with Tor to help prepare the user research for Phase 7 of Sponsor 9 until early February:
https://gitlab.torproject.org/tpo/ux/research/-/issues/128
In December, I did an inventory of user research techniques that could be well suited to the limitations and opportunities of outsourcing user research during trainings and discussed them with the UX team:
- https://gitlab.torproject.org/tpo/ux/research/-/work_items/129
Then, I explored a bit more our top 3 candidate research activities:
- I brainstormed on how a diary study of Tor Browser could work:
https://gitlab.torproject.org/tpo/ux/research/-/work_items/132
- I prototyped a desirability study of @nicob's new illustrations:
https://gitlab.torproject.org/tpo/ux/research/-/work_items/133
- I prototyped a card sorting to reorganize support.torproject.org:
https://gitlab.torproject.org/tpo/ux/research/-/issues/8
I also researched how to create a demographics questionnaire that people call fill up either online or offline and that make data gathering and reporting easier for us. We might use either a PDF form or LimeSurvey:
https://gitlab.torproject.org/tpo/ux/research/-/work_items/130
The next step on both fronts is to decide with the UX Team on which 2 user research activities to pursue and which technology to use for the demographics questionnaire.
In January, I almost finished preparing the 2 user research activities for Phase 7 of Sponsor 9.
We chose 2 activites that:
- Don't require much facilitation skills from partners - Are lightweight for partners to report - Benefit from a slightly bigger numbers of participants (~20) - Are easy to schedule during the training
We decided to organize:
- Desirability study of the new illustrations
https://gitlab.torproject.org/tpo/ux/research/-/work_items/133
This one takes the form of a survey that all training participants can fill up on their own at the beginning of the training.
https://survey.torproject.org/index.php/983521
- Card sorting on support portal
Partners can get several participants to do the card sorting in parallel and ask them to self-report their results as a survey.
https://survey.torproject.org/index.php/168241
As you can see in both, I also implemented a reusable consent and demographics questionnaire that can be either added at the beginning of another research survey or printed on paper.
I created a Tor theme for LimeSurvey that uses the Tor colors and logo, prints well on paper, and displays better on mobile. Thanks @lavamind for the speedy deployment!
https://gitlab.torproject.org/tpo/web/fruity_twentythree_tor
Today is the last day of work on my contract for Sponsor 09, so I'm already sending my report for February.
In February,
I finished preparing the 2 user research activities for Phase 7 of Sponsor 9. You can find the preparation material for both studies on GitLab.
- Desirability study of the new illustrations
https://gitlab.torproject.org/tpo/ux/research/-/work_items/133
- Card sorting on support portal
https://gitlab.torproject.org/tpo/ux/research/-/issues/8
I also prepared a merge request to publish a data handling policy for user research data on community.torproject.org:
https://gitlab.torproject.org/tpo/ux/research/-/issues/90
Cheers,
End of March, I started another short contract with Tor to help with:
- S09: Analysis of the user research from the Privacy Resilience Grants - S131: Mullvad Browser system installer - S101: Usability testing of Tor VPN
In April,
- S09
Finished preparing the material for the user research and onboarded the 3 partners who are going to do the research in East Africa in April-May.
* Desirability study of new illustrations: https://gitlab.torproject.org/tpo/ux/research/-/issues/137
* Card sorting on support-dot-menu: https://gitlab.torproject.org/tpo/ux/research/-/issues/8
- S131
Helped @pierov and @ruihildt with their discussion on the Mullvad Browser installer and drew 2 iterations of mockups for the whole flow.
https://gitlab.torproject.org/tpo/applications/mullvad-browser/-/issues/200
- S101
Discussed a timeline and scope for the usability tests and drafted tasks and interview questions.
https://gitlab.torproject.org/tpo/ux/research/-/issues/69
The tests are scheduled for May 20-26 in Mexico City, while you're all busy in Portugal :)
Prepared a screening survey and started recruiting for the tests.
Tested 0.5.0 and 0.6.0 and reported 3 important bugs that should be fixed before we do the usability tests.
* https://gitlab.torproject.org/tpo/applications/vpn/-/issues/158 * https://gitlab.torproject.org/tpo/applications/vpn/-/issues/159 * https://gitlab.torproject.org/tpo/applications/vpn/-/issues/160
Cheers,
Here is what I did in May on S09, S101, and S131.
### Desirability study (S09, ux/research#137)
We received 54 answers to the survey for trainees who took part of the Privacy Resilience Grants in East Africa. That's lots of data. Great!
I built a spreadsheet to analyze all this data.
### Card sorting (S09, ux/research#8)
We received 18 results from the card sorting facilitated by the 3 partners in East Africa. Unfortunately, I discarded 11 of these results because of the very poor quality of the data. That left me with too little data to do a solid correlation analysis between the sorts.
Thankfully, @SugarArchivist saved the day by conducting 4 more sorts in the US. That's still not a lot for strong correlation, but that gives me enough material to do some analysis.
I started doing the analysis using a methodology by Donna Spencer:
https://rosenfeldmedia.com/people/donna-spencer/
### Mullvad Browser installer (S131, applications/mullvad-browser#200)
Helped @pierov and @ruihildt polish the copy for the Mullvad Browser installer. We're almost there!
### Usability tests of Tor VPN (S101, ux/research#69)
I tested Tor VPN some more to understand better when bridges don't work.
https://gitlab.torproject.org/tpo/core/onionmasq/-/issues/97
The tests are ready to be run whenever we have a good enough prototype!
- I identified 10 strong candidates for the usability tests who already used VPNs and some of them Tor as well.
- I prepared realistic tasks for them to perform on Tor VPN and aligned with the priorities identified with the development team:
- I redacted stepped tasks (in Spanish):
https://gitlab.torproject.org/tpo/ux/research/-/issues/69#user-tasks
- I improved my network to simulate censorship of Telegram and websites.
- I prepared the consent paperwork for the tests.
My second short contract this year ended last week, so I'm reporting for June already.
### Desirability study (S09, ux/research#137)
Finished analyzing the desirability study that we did as part of the Privacy Resilience Grants with partners in East Africa.
Report: https://gitlab.torproject.org/tpo/ux/research/uploads/f64496059f692f7d071355...
The new illustrations were described positively by participants overall.
Tor was perceived as advanced technology that is helpful, effective, and trustworthy, but not as approachable as we want to be.
To improve, we should focus on simplicity and clarity.
### Card sorting (S09, ux/research#8)
Finished analyzing the card sorting that we also did as part of the Privacy Resilience Grants.
Report: https://gitlab.torproject.org/tpo/ux/research/uploads/1cf2bd5006067633698ac3...
Most participants used a classification scheme based on value (or threat, it's opposite) with categories such as "Censorship", "Privacy", "Identity", and "Connectivity", while support.torproject.org has much more emphasis on technologies, with categories such as "Tor Browser", "Tor Mobile", "HTTPS", etc.
### Retrospective on user research for Sponsor 9
I did a solo retrospective on these 2 studies and Sponsor 9 in general.
For both studies, I think that we would have had better results, in less time, and maybe even cheaper, by doing the studies ourselves.
https://gitlab.torproject.org/tpo/ux/research/-/issues/138
### Mullvad Browser installer (S131, applications/mullvad-browser#200)
Reviewed the strings for Mullvad Browser Installer:
https://gitlab.torproject.org/tpo/applications/tor-browser-build/-/merge_req...
tor-project@lists.torproject.org